Effects of elevated CO 2 and O 3 on leaf litter phenolics and subsequent performance of litter-feeding soil macrofauna

Two field-growing silver birch (Betula pendula Roth) clones (clone 4 and 80) were exposed to elevated CO2 and O3 for three growing seasons (1999–2001). The phenolic compounds of naturally abscised leaf litter were analyzed in order to determine the possible CO2- and O3-induced changes in the litter quality. The potential litter-mediated CO2 and O3 effects on litter-feeding soil macrofauna (detritivore) performance were assessed in microcosm experiments, i.e., the relative growth rates (RGR) of Lumbricus terrestris and Porcellio scaber, the relative consumption rates (RCR) of P. scaber, and mortality of the test animals were measured. The leaf litter grown under elevated CO2 had increased concentrations (weight per mass unit) and contents (weight per leaf) of phenolic acids, flavonol glycosides, condensed tannins and total measured phenolics. Elevated O3 increased the concentrations of 3,4'-dihydroxypropiophenone 3-β-D-glucoside (DHPPG) and flavonoid aglycones but only under ambient CO2. However, elevated O3 effects on the content of some low-molecular-weight phenolic (LMWP) compounds (i.e. phenolic acids, DHPPG, flavonoid aglycones) and total LMWP changed over time emphasizing the importance of conducting long-term (>3 years) exposure studies. In general, RGR of young L. terrestris was affected by the litter quality changes induced by elevated CO2 and O3, as the animal growth rates were reduced when they were fed with CO2- and O3-exposed leaf litter of clone 80 in Experiment 1. P. scaber RCR or RGR responses to CO2- and O3-induced changes in litter quality were more variable and inconsistent, and neither were there any litter-mediated CO2 and O3 effects on animal mortality in these microcosm experiments. In conclusion, elevated CO2 has the potential to alter silver birch leaf litter quality, but the possible O3 effects on phenolic compounds and litter-mediated CO2 and O3 effects on detritivores are more difficult to validate.